added contours compare, hist compare, knearest (OCR), cleanups

svn path=/trunk/externals/pdp_opencv/; revision=11920

1 files changed, 552 insertions, 0 deletions

diff --git a/pdp_opencv_knear.cc b/pdp_opencv_knear.cc
new file mode 100755
index 0000000..c94d7e4
--- /dev/null
+++ b/pdp_opencv_knear.cc
@@ -0,0 +1,552 @@
+/*
+ *   Pure Data Packet module.
+ *   Copyright (c) by Tom Schouten <pdp@zzz.kotnet.org>
+ *
+ *   This program is free software; you can redistribute it and/or modify
+ *   it under the terms of the GNU General Public License as published by
+ *   the Free Software Foundation; either version 2 of the License, or
+ *   (at your option) any later version.
+ *
+ *   This program is distributed in the hope that it will be useful,
+ *   but WITHOUT ANY WARRANTY; without even the implied warranty of
+ *   MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ *   GNU General Public License for more details.
+ *
+ *   You should have received a copy of the GNU General Public License
+ *   along with this program; if not, write to the Free Software
+ *   Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.
+ *
+ *
+ *   pdp_opencv_knear : OCR like pattern recognition
+ *   based on basic OCR with Open CV tutorial
+ *   by damiles : http://blog.damiles.com/?p=93
+ *
+ *
+ */
+
+#include <stdio.h>
+#include <stdlib.h>
+#include <string.h>
+#include <dirent.h>
+#include <limits.h>
+#include <dlfcn.h>
+#include <ctype.h>
+
+#include "pdp.h"
+
+#include "cv.h"
+#include "highgui.h"
+#include "ml.h"
+
+typedef struct pdp_opencv_knear_struct
+{
+  t_object x_obj;
+  t_float x_f;
+
+  t_outlet *x_outlet0;
+  t_outlet *x_outlet1;
+
+  int x_packet0;
+  int x_packet1;
+  int x_dropped;
+  int x_queue_id;
+
+  int x_width;
+  int x_height;
+  int x_size;
+
+  int  x_classify;
+  int  x_ROIx;
+  int  x_ROIy;
+  int  x_ROIw;
+  int  x_ROIh;
+
+  // The output and temporary images
+  IplImage *rgb, *grey;
+
+  // open cv classifier data
+  char *x_filepath;
+  int  x_nsamples;
+  int  x_rsamples;
+  CvMat* trainData;
+  CvMat* trainClasses;
+  CvMat* x_nearest;
+  CvMat* x_dist;
+  int x_pwidth;
+  int x_pheight;
+  CvKNearest *knn;
+
+} t_pdp_opencv_knear;
+
+IplImage pdp_opencv_knear_preprocessing(IplImage* imgSrc,int new_width, int new_height);
+
+static void pdp_opencv_knear_process_rgb(t_pdp_opencv_knear *x)
+{
+    t_pdp     *header = pdp_packet_header(x->x_packet0);
+    short int *data   = (short int *)pdp_packet_data(x->x_packet0);
+    t_pdp     *newheader = pdp_packet_header(x->x_packet1);
+    short int *newdata = (short int *)pdp_packet_data(x->x_packet1); 
+    int i,j,k,im;
+
+    if ((x->x_width != (t_int)header->info.image.width) || 
+        (x->x_height != (t_int)header->info.image.height)) 
+    {
+
+      post("pdp_opencv_knear :: resizing");
+  
+      x->x_width = header->info.image.width;
+      x->x_height = header->info.image.height;
+      x->x_size = x->x_width*x->x_height;
+      x->x_ROIx = 0;
+      x->x_ROIy = 0;
+      x->x_ROIw = header->info.image.width;
+      x->x_ROIh = header->info.image.height;
+
+      cvReleaseImage( &x->rgb );
+      cvReleaseImage( &x->grey );
+
+      // create cv_images
+      x->rgb = cvCreateImage( cvSize(x->x_width, x->x_height), 8, 3 );
+      x->grey = cvCreateImage( cvSize(x->x_width, x->x_height), 8, 1 );
+    
+    }
+    
+    newheader->info.image.encoding = header->info.image.encoding;
+    newheader->info.image.width = x->x_width;
+    newheader->info.image.height = x->x_height;
+
+    memcpy( x->rgb->imageData, data, x->x_size*3 );
+        
+    cvCvtColor( x->rgb, x->grey, CV_BGR2GRAY );
+
+    if ( x->x_classify )
+    {
+     IplImage prs_image;
+     float result;
+     CvMat row_header, *row1, odata;
+
+       // post( "pdp_opencv_knear : size : (%dx%d)", x->x_pwidth, x->x_pheight);
+
+       // process file
+       prs_image = pdp_opencv_knear_preprocessing(x->grey, x->x_pwidth, x->x_pheight);
+
+       //Set data
+       IplImage* img32 = cvCreateImage( cvSize( x->x_pwidth, x->x_pheight ), IPL_DEPTH_32F, 1 );
+       cvConvertScale(&prs_image, img32, 0.0039215, 0);
+       cvGetSubRect(img32, &odata, cvRect(0,0, x->x_pwidth, x->x_pheight));
+       row1 = cvReshape( &odata, &row_header, 0, 1 );
+
+       result=x->knn->find_nearest(row1,x->x_nsamples,0,0,x->x_nearest,x->x_dist);
+       // for ( i=0; i<x->x_nsamples; i++ )
+       // {
+         // post( "pdp_opencv_knear : distance : %f", x->x_dist->data.fl[i] );
+       // }
+       outlet_float(x->x_outlet1, x->x_dist->data.fl[0]);
+
+       cvReleaseImage( &img32 );
+       x->x_classify = 0;
+    }
+
+    memcpy( newdata, x->rgb->imageData, x->x_size*3 );
+    return;
+}
+
+static void pdp_opencv_knear_sendpacket(t_pdp_opencv_knear *x)
+{
+  /* release the packet */
+  pdp_packet_mark_unused(x->x_packet0);
+  x->x_packet0 = -1;
+
+  /* unregister and propagate if valid dest packet */
+  pdp_packet_pass_if_valid(x->x_outlet0, &x->x_packet1);
+}
+
+static void pdp_opencv_knear_process(t_pdp_opencv_knear *x)
+{
+   int encoding;
+   t_pdp *header = 0;
+   char *parname;
+   unsigned pi;
+   int partype;
+   float pardefault;
+   t_atom plist[2];
+   t_atom tlist[2];
+   t_atom vlist[2];
+
+  /* check if image data packets are compatible */
+  if ( (header = pdp_packet_header(x->x_packet0))
+  && (PDP_BITMAP == header->type)){
+    
+  /* pdp_opencv_knear_process inputs and write into active inlet */
+  switch(pdp_packet_header(x->x_packet0)->info.image.encoding){
+
+   case PDP_BITMAP_RGB:
+            x->x_packet1 = pdp_packet_clone_rw(x->x_packet0);
+            pdp_queue_add(x, (void*)pdp_opencv_knear_process_rgb, (void*)pdp_opencv_knear_sendpacket, &x->x_queue_id);
+      break;
+
+   default:
+      /* don't know the type, so dont pdp_opencv_knear_process */
+      break;
+      
+   }
+  }
+
+}
+
+static void pdp_opencv_knear_input_0(t_pdp_opencv_knear *x, t_symbol *s, t_floatarg f)
+{
+    /* if this is a register_ro message or register_rw message, register with packet factory */
+
+    if (s == gensym("register_rw")) 
+       x->x_dropped = pdp_packet_convert_ro_or_drop(&x->x_packet0, (int)f, pdp_gensym((char*)"bitmap/rgb/*") );
+
+    if ((s == gensym("process")) && (-1 != x->x_packet0) && (!x->x_dropped))
+    {
+        /* add the process method and callback to the process queue */
+        pdp_opencv_knear_process(x);
+    }
+}
+
+static void pdp_opencv_knear_free(t_pdp_opencv_knear *x)
+{
+  int i;
+
+    pdp_queue_finish(x->x_queue_id);
+    pdp_packet_mark_unused(x->x_packet0);
+
+    // destroy cv_images
+    cvReleaseImage( &x->rgb );
+    cvReleaseImage( &x->grey );
+    cvReleaseMat( &x->trainData );
+    cvReleaseMat( &x->trainClasses );
+
+}
+
+t_class *pdp_opencv_knear_class;
+
+//
+//
+// Find the min box. The min box respect original aspect ratio image
+// The image is a binary data and background is white.
+//
+//
+
+
+void pdp_opencv_knear_findX(IplImage* imgSrc,int* min, int* max)
+{
+  int i;
+  int minFound=0;
+  CvMat data;
+  CvScalar maxVal=cvRealScalar(imgSrc->width * 255);
+  CvScalar val=cvRealScalar(0);
+
+  // for each col sum, if sum < width*255 then we find the min
+  // then continue to end to search the max, if sum< width*255 then is new max
+  for (i=0; i< imgSrc->width; i++)
+  {
+    cvGetCol(imgSrc, &data, i);
+    val= cvSum(&data);
+    if(val.val[0] < maxVal.val[0])
+    {
+      *max= i;
+      if(!minFound)
+      {
+        *min= i;
+         minFound= 1;
+      }
+    }
+  }
+}
+
+void pdp_opencv_knear_findY(IplImage* imgSrc,int* min, int* max)
+{
+  int i;
+  int minFound=0;
+  CvMat data;
+  CvScalar maxVal=cvRealScalar(imgSrc->width * 255);
+  CvScalar val=cvRealScalar(0);
+
+  // for each col sum, if sum < width*255 then we find the min
+  // then continue to end to search the max, if sum< width*255 then is new max
+  for (i=0; i< imgSrc->height; i++)
+  {
+    cvGetRow(imgSrc, &data, i);
+    val= cvSum(&data);
+    if(val.val[0] < maxVal.val[0])
+    {
+      *max=i;
+      if(!minFound)
+      {
+        *min= i;
+         minFound= 1;
+      }
+    }
+  }
+}
+
+//
+//
+// Find the bounding box. 
+//
+//
+
+CvRect pdp_opencv_knear_findBB(IplImage* imgSrc)
+{
+  CvRect aux;
+  int xmin, xmax, ymin, ymax;
+  xmin=xmax=ymin=ymax=0;
+
+  pdp_opencv_knear_findX(imgSrc, &xmin, &xmax);
+  pdp_opencv_knear_findY(imgSrc, &ymin, &ymax);
+
+  aux=cvRect(xmin, ymin, xmax-xmin, ymax-ymin);
+
+  return aux;
+}
+
+IplImage pdp_opencv_knear_preprocessing(IplImage* imgSrc,int new_width, int new_height)
+{
+  IplImage* result;
+  IplImage* scaledResult;
+
+  CvMat data;
+  CvMat dataA;
+  CvRect bb;//bounding box
+  CvRect bba;//boundinb box maintain aspect ratio
+
+  // find bounding box
+  bb=pdp_opencv_knear_findBB(imgSrc);
+
+  if ( ( bb.width == 0 ) || ( bb.height == 0 ) )
+  {
+     bb.x = 0;
+     bb.y = 0;
+     bb.width = imgSrc->width;
+     bb.height = imgSrc->height;
+  }
+
+  // get bounding box data and no with aspect ratio, the x and y can be corrupted
+  cvGetSubRect(imgSrc, &data, cvRect(bb.x, bb.y, bb.width, bb.height));
+  // create image with this data with width and height with aspect ratio 1
+  // then we get highest size betwen width and height of our bounding box
+  int size=(bb.width>bb.height)?bb.width:bb.height;
+  result=cvCreateImage( cvSize( size, size ), 8, 1 );
+  cvSet(result,CV_RGB(255,255,255),NULL);
+  // copy de data in center of image
+  int x=(int)floor((float)(size-bb.width)/2.0f);
+  int y=(int)floor((float)(size-bb.height)/2.0f);
+  cvGetSubRect(result, &dataA, cvRect(x,y,bb.width, bb.height));
+  cvCopy(&data, &dataA, NULL);
+  // scale result
+  scaledResult=cvCreateImage( cvSize( new_width, new_height ), 8, 1 );
+  cvResize(result, scaledResult, CV_INTER_NN);
+
+  // return processed data
+  return *scaledResult;
+
+}
+
+void pdp_opencv_knear_load(t_pdp_opencv_knear *x)
+{
+  IplImage* src_image;
+  IplImage prs_image;
+  CvMat row,data;
+  char file[255];
+  int i=0,j;
+  CvMat row_header, *row1;
+
+  x->x_rsamples = 0;
+  
+  for( j = 0; j< x->x_nsamples; j++)
+  {
+
+     // load file
+     sprintf(file,"%s/%03d.png",x->x_filepath, j);
+     src_image = cvLoadImage(file,0);
+     if(!src_image)
+     {
+       post("pdp_opencv_knear : error: couldn't load image %s\n", file);
+       continue;
+     }
+     if ( ( x->x_pwidth == -1 ) || ( x->x_pheight == -1 ) )
+     {
+        x->x_pwidth = src_image->width;
+        x->x_pheight = src_image->height;
+        // post( "pdp_opencv_knear : loaded : %s (%dx%d)", file, src_image->width, src_image->height);
+        x->x_rsamples++;
+     }
+     else if ( ( src_image->width != x->x_pwidth ) || ( src_image->height != x->x_pheight ) )
+     {
+        post( "pdp_opencv_knear : error : %s (%dx%d) : wrong size ( should be %dx%d )", file, src_image->width, src_image->height, x->x_pwidth, x->x_pheight);
+        continue;
+     }
+     else
+     {
+        // post( "pdp_opencv_knear : loaded : %s (%dx%d)", file, src_image->width, src_image->height);
+        x->x_rsamples++;
+     }
+
+     // process file
+     prs_image = pdp_opencv_knear_preprocessing(src_image, x->x_pwidth, x->x_pheight);
+     // post( "pdp_opencv_knear : preprocessed : %s (%dx%d)", file, x->x_pwidth, x->x_pheight);
+
+     if ( ( x->trainData == NULL ) || ( x->trainClasses == NULL ))
+     {
+       x->trainData = cvCreateMat(x->x_nsamples, x->x_pwidth*x->x_pheight, CV_32FC1);
+       x->trainClasses = cvCreateMat(x->x_nsamples, 1, CV_32FC1);
+     }
+
+     // set class label
+     cvGetRow(x->trainClasses, &row, j);
+     cvSet(&row, cvRealScalar(i), NULL);
+     // set data
+     cvGetRow(x->trainData, &row, j);
+
+     IplImage* img = cvCreateImage( cvSize( x->x_pwidth, x->x_pheight ), IPL_DEPTH_32F, 1 );
+     // convert 8 bits image to 32 float image
+     cvConvertScale(&prs_image, img, 0.0039215, 0);
+
+     cvGetSubRect(img, &data, cvRect( 0, 0, x->x_pwidth, x->x_pheight) );
+
+     // convert data matrix sizexsize to vecor
+     row1 = cvReshape( &data, &row_header, 0, 1 );
+     cvCopy(row1, &row, NULL);
+     cvReleaseImage( &img );
+  }
+
+  // create the classifier
+  post( "pdp_opencv_knear : loaded : %d samples", x->x_rsamples);
+  if ( x->x_rsamples == x->x_nsamples )
+  {
+    x->knn=new CvKNearest( x->trainData, x->trainClasses, 0, false, x->x_nsamples );
+    x->x_nearest=cvCreateMat(1,x->x_nsamples,CV_32FC1);
+    x->x_dist=cvCreateMat(1,x->x_nsamples,CV_32FC1);
+  }
+}
+
+static void pdp_opencv_knear_classify(t_pdp_opencv_knear *x)
+{
+    if ( x->trainData == NULL )
+    {
+       post( "pdp_opencv_knear : no patterns loaded : cannot process" );
+       return;
+    }
+    x->x_classify=1;
+}
+
+static void pdp_opencv_knear_pload(t_pdp_opencv_knear *x, t_symbol *path, t_floatarg nsamples )
+{
+   if ( (int) nsamples <= 0 )
+   {
+      post( "pdp_opencv_knear : wrong number of samples : %d", nsamples );
+      return;
+   }
+   else
+   {
+      x->x_nsamples = (int)nsamples;
+      x->x_rsamples = 0;
+      cvReleaseMat( &x->trainData );
+      cvReleaseMat( &x->trainClasses );
+      x->trainData = NULL;
+      x->trainClasses = NULL;
+   }
+   strcpy( x->x_filepath, path->s_name );
+   pdp_opencv_knear_load(x);
+}
+
+void *pdp_opencv_knear_new(t_symbol *s, int argc, t_atom *argv )
+{
+  int i;
+
+  t_pdp_opencv_knear *x = (t_pdp_opencv_knear *)pd_new(pdp_opencv_knear_class);
+
+  x->x_outlet0 = outlet_new(&x->x_obj, &s_anything); 
+  x->x_outlet1 = outlet_new(&x->x_obj, &s_anything);
+
+  x->x_packet0 = -1;
+  x->x_packet1 = -1;
+  x->x_queue_id = -1;
+
+  x->x_width  = 320;
+  x->x_height = 240;
+  x->x_size   = x->x_width * x->x_height;
+  x->x_ROIx = 0;
+  x->x_ROIy = 0;
+  x->x_ROIw = x->x_width;
+  x->x_ROIh = x->x_height;
+
+  x->rgb = cvCreateImage( cvSize(x->x_width, x->x_height), 8, 3 );
+  x->grey = cvCreateImage( cvSize(x->x_width, x->x_height), 8, 1 );
+
+  x->x_filepath = ( char * ) getbytes( 1024 );
+  sprintf( x->x_filepath, "./data" );
+  x->x_nsamples = 10;
+
+  if ( argc >= 1 )
+  {
+      if ( argv[0].a_type != A_SYMBOL )
+      {
+        error( "pdp_opencv_knear : wrong argument (file path : 1)" );
+        return NULL;
+      }
+      if ( !strcmp( argv[0].a_w.w_symbol->s_name, "" ) )
+      {
+        error( "pdp_opencv_knear : no extension specified" );
+        error( "pdp_opencv_knear : usage : pdp_opencv_knear_new <file path> <nsamples>" );
+        return NULL;
+      }
+      strcpy( x->x_filepath, argv[0].a_w.w_symbol->s_name );
+  }
+  if ( argc >= 2 )
+  {
+      if ( argv[1].a_type != A_FLOAT )
+      {
+        error( "pdp_opencv_knear : wrong argument (nsamples : 2)" );
+        return NULL;
+      }
+      if ( (int)argv[1].a_w.w_float <= 0 )
+      {
+        error( "pdp_opencv_knear : wrong number of samples (%d)", (t_int)(int)argv[1].a_w.w_float );
+        error( "pdp_opencv_knear : usage : pdp_opencv_knear_new <file path> <nsamples>" );
+        return NULL;
+      }
+      x->x_nsamples = (int)argv[1].a_w.w_float;
+  }
+
+  x->x_classify = 0;
+  x->x_pwidth = -1;
+  x->x_pheight = -1;
+
+  x->trainData = NULL;
+  x->trainClasses = NULL;
+
+  post( "pdp_opencv_knear : loading %d samples from : %s", x->x_nsamples, x->x_filepath );
+  pdp_opencv_knear_load( x );
+
+  return (void *)x;
+}
+
+
+#ifdef __cplusplus
+extern "C"
+{
+#endif
+
+
+void pdp_opencv_knear_setup(void)
+{
+
+    post( "    pdp_opencv_knear");
+    pdp_opencv_knear_class = class_new(gensym("pdp_opencv_knear"), (t_newmethod)pdp_opencv_knear_new,
+      (t_method)pdp_opencv_knear_free, sizeof(t_pdp_opencv_knear), CLASS_DEFAULT, A_GIMME, A_NULL);
+
+    class_addmethod(pdp_opencv_knear_class, (t_method)pdp_opencv_knear_input_0, gensym("pdp"), A_SYMBOL, A_DEFFLOAT, A_NULL);
+    class_addmethod(pdp_opencv_knear_class, (t_method)pdp_opencv_knear_classify, gensym("bang"), A_NULL);
+    class_addmethod(pdp_opencv_knear_class, (t_method)pdp_opencv_knear_pload, gensym("load"), A_SYMBOL, A_DEFFLOAT, A_NULL);
+
+}
+
+#ifdef __cplusplus
+}
+#endif